Am. J. Hum. Genet.

66:14961503, 2000

Neurological Phenotype in Waardenburg Syndrome Type 4

Correlates with Novel SOX10 Truncating Mutations
and Expression in Developing Brain
Renaud L. Touraine,1,* Tania Attie-Bitach,1,* Eric Manceau,2 Eckhard Korsch,3 Pierre Sarda,5
Veronique Pingault,6 Ferechte Encha-Razavi,1 Anna Pelet,1 Joelle Auge,1
Annie Nivelon-Chevallier,2 Alexander Mathias Holschneider,3 Marc Munnes,4 Walter Doerfler,4
Michel Goossens,5 Arnold Munnich,1 Michel Vekemans,1 and Stanislas Lyonnet1
1
Departement de Genetique et Unite de Recherches sur les Handicaps Genetiques de lEnfant, Hopital Necker-Enfants Malades, Paris;
2
Centre de Genetique, Hopital dEnfants, Dijon, France; 3Kinderkrankenhaus, Kliniken der Stadt Koln, and 4Institut fur Genetik, Universitat zu
Koln, Cologne, Germany; 5Unite de Genetique Medicale et Foetopathologie, Hopital Arnaud de Villeneuve, Montpellier, France; and
6
Genetique Moleculaire et Physiopathologie, Hopital Henri Mondor, Creteil, France

Waardenburg syndrome type 4 (WS4), also called Shah-Waardenburg syndrome, is a rare neurocristopathy that
results from the absence of melanocytes and intrinsic ganglion cells of the terminal hindgut. WS4 is inherited as
an autosomal recessive trait attributable to EDN3 or EDNRB mutations. It is inherited as an autosomal dominant
condition when SOX10 mutations are involved. We report on three unrelated WS4 patients with growth retardation
and an as-yet-unreported neurological phenotype with impairment of both the central and autonomous nervous
systems and occasionally neonatal hypotonia and arthrogryposis. Each of the three patients was heterozygous for
a SOX10 truncating mutation (Y313X in two patients and S351X in one patient). The extended spectrum of the
WS4 phenotype is relevant to the brain expression of SOX10 during human embryonic and fetal development.
Indeed, the expression of SOX10 in human embryo was not restricted to neural-crestderived cells but also involved
fetal brain cells, most likely of glial origin. These data emphasize the important role of SOX10 in early development
of both neural-crestderived tissues, namely melanocytes, autonomic and enteric nervous systems, and glial cells
of the central nervous system.

Introduction features reminiscent of WS, such as hypomelanotic

Waardenburg syndrome (WS [MIM 193500, MIM
193510, MIM 600193, and MIM 148820]) is a rare
disorder (1/50,000 live births) characterized by sen-
sorineural hearing loss and pigment anomalies. On
the other hand, Hirschsprung disease (HSCR [MIM
142623]) is a common malformation (1/5,000 live
births) defined by the absence of parasympathetic gan-
glion cells in the terminal hindgut. HSCR and WS are
regarded as neurocristopathies, since both disorders re-
sult from an abnormal migration of neural-crestderived
cells. Interestingly, each of the three animal models of
HSCR, namely the piebald lethal (sl), lethal spotting (ls),
and dominant megacolon (Dom) mutants, exhibit some
Received June 22, 1999; accepted for publication February 16, 2000;
electronically published April 4, 2000.
Address for correspondence and reprints: Stanislas Lyonnet, De-
partement de Genetique et Unite de Recherches sur les Handicaps
Genetiques de lEnfant, INSERM U-393, Hopital Necker-Enfants Ma-
lades, 75743 Paris. E-mail: lyonnet@necker.fr
*
These two authors equally contributed to this work.
q 2000 by The American Society of Human Genetics. All rights reserved.
0002-9297/2000/6605-0004$02.00
spots. In humans, the WS-HSCR association (WS type
4 [WS4], also known as Shah-Waardenburg syndrome;
MIM 277580) is a very rare and heterogeneous condi-
tion that follows either an autosomal recessive or an
autosomal dominant pattern of inheritance (McKusick
1973; Branski et al. 1979; Omenn and McKusick 1979;
Shah et al. 1981). Both the mouse mutants and the hu-
man WS4 phenotype have been ascribed to mutations
of the endothelin-signaling pathwaynamely, the en-
dothelin B receptor, EDNRB, and the endothelin 3 gene,
EDN3 (Attie et al. 1995; Edery et al. 1996)or of the
SOX10 gene, which encodes a transcription factor (Pin-
gault et al. 1998). Hitherto, no additional features have
been described among WS4 patients. Here, we report
on three unrelated WS4 patients who presented with an
as-yet-unreported progressive neurological involvement
of both central and peripheral nervous systems. We show
that the three patients carried a truncating SOX10 mu-
tation. Accordingly, analysis of data generated from in
situ hybridization experiments indicated that SOX10 is
expressed early in the fetal brain, likely in glial cells.
Thus, our study gives support to a possible neurological

1496
Touraine et al.: SOX10 Mutation in a Neurological Variant of WS4 1497

involvement in WS4 and to the role of SOX10 in de- Table 1

velopment of both the central and the autonomous nerv- WS4 Phenotype for Members of the Three Families Studied
ous systems.
PRESENCE OF FEATURE IN

FEATURE Family 1 Family 2 Family 3a

Patients and Methods
Colonic aganglionosis Total Short segment Total
Iris heterochromia 1 1 2
Patients and Families Deafness 1 1 1
Mental retardation Severe Very mild ND
Histopathologic inclusion criteria for HSCR were gan- Cerebellar ataxia 1 1 ND
glia that were absent, along with increased acetylcho- Nystagmus 1 1 ND
linesterase histochemical staining in nerve fibers on suc- Spasticity 1 1 1
tion biopsies of the rectal mucosa. For WS, we used the Alacrima 1 1 ND
diagnostic criteria of Read and Newton (1997), and Asialia 2 1 ND
Reduced sweating 2 1 ND
WS4 was diagnosed when at least one member of the Seizures 2 1 2
probands family had features of both HSCR and WS. Growth failure 1 1 ND
Informed consent was obtained for all the patients in Hypogonadism 1 2 ND
the study. NOTE.A plus sign (1) indicates presence, and a minus sign
Among a total of 12 unrelated WS4 patients recruited (2) indicates absence.
by the Genetic Center, Necker Hospital, Pais, three pre- a
ND = not determined.
sented with additional neurological symptoms. Each of
them originated from white families living in Germany limbs without accompanying abnormalities (as assessed
(family 1) and France (families 2 and 3). Their standard by electroencephalogram or brain magnetic-resonance
blood-lymphocyte karyotypes were within normal ref- imaging). On the basis of a later occurrence of torticollis
erence ranges. and the benign course of the nystagmus, head bobbing,
Patient 1 had early delayed developmental milestones, and paroxysmal movement, a putative diagnosis of spas-
nystagmus, myopia, reduced tear production, hypo- mus nutans was made, despite the early onset of symp-
tonia, and growth deficiency (height, 104 cm; weight, toms and despite the presence of abnormal movements
15.8 kg; and head circumference, 52 cm at age 12 years). of the limbs. Later, antiepileptic drugs were withdrawn
In addition to extensive intestinal aganglionosis, deaf- without any recurrence of seizures. Ataxia, spasticity,
ness, iris heterochromia, and hypomelanic skin patches and reduced productions of tears, saliva, and sweat de-
consistent with WS4, patient 1 developed cerebellar veloped progressively. Growth failure occurred at age 8
ataxia, spasticity, and severe mental retardation (table years, and, at the time of the study, he was on the 22
1). Several computed-tomography scans revealed only a SD curve. He had neither hypomelanic lesions nor he-
slight cerebral atrophy. Pupillary light reflexes, optical patosplenomegaly. Neither his parents nor any other
fundus, sensitivity to pain and temperature, sense of family member had any signs of WS or HSCR.
smell, sweating, and salivation were normal. Echocar- The proband of family 3 was the first child of young
diography and 24-h electrocardiogram were unremark- parents. There was no history of either WS- or HSCR-
able, but heart-rate variation parameters indicated an affected members of the probands family. A reduction
autonomic dysregulation. Similarly, there was almost no of fetal movement was noted during the last month of
local reaction to the histamine test (flare !2 mm). Fur- pregnancy. Height, weight, and head circumference at
thermore, he developed hepatosplenomegaly and portal birth were 50 cm, 2,980 g, and 35.5 cm, respectively.
hypertension of unknown origin. Three other boys who Severe neonatal distress required immediate resuscita-
were members of his family had died during the first tion and admission to an intensive care unit. He pre-
week of life; they had intestinal obstruction, and it was sented with coma, arthrogryposis, meconial ileus, and a
likely that they had WS, since the third child had patchy white forelock, but he did not have heterochromia irides.
depigmentation of the scalp and white hair (fig. 1). Nei- Analysis of intestinal biopsies led to the diagnosis of
ther the mother nor any other family member had any severe aganglionosis extending to the jejunum. Optical
signs of WS or HSCR. fundus were normal. Neither auditory brain stem
In family 2, the proband was initially diagnosed with evoked responses nor electromyographic reactivity could
short-segment aganglionosis; he then presented with be found. Electroencephalogram was abnormal, show-
myoclonia at age 2 wk. Computed-tomography scans of ing immaturity and discontinuity of the tracing, but par-
his brain revealed no abnormalities. He responded well oxysmal discharges were not present. Analysis of muscle
to phenobarbitone, but, later, he developed nystagmus, biopsy showed no specific lesions. He died on day 11
head bobbing, and paroxysmal movements of the upper of life.
1498 Am. J. Hum. Genet. 66:14961503, 2000

Figure 1 Y313X and S251X SOX10 mutations. A, DNA sequence of SOX10 exon 5 showing the normal sequence (N) and the Y313X
heterozygous mutation (arrow) in the proband (family 1). B, DNA sequence of SOX10 exon 5 showing the normal sequence (N) and the S251X
heterozygous mutation in the proband (family 3). C, Family 1 with pedigree and PCR-amplified exon 5 of the SOX10 gene, digested with DdeI:
normal allele (1) = 201 1 66 bp (132-bp fragment nonvisible), mutated allele (m) = 256 bp (132-bp fragment nonvisible).

DNA Analysis antisense riboprobes were generated by either T7 or T3

Mutation screening of SOX10 coding sequence was
performed by SSCP as previously described, except gel
electrophoreses were performed in two different con-
ditions: at room temperature and in a cold room (Pin-
gault et al. 1998). In brief, PCR reactions were per-
formed on 200 ng of DNA with Taq DNA polymerase
(BRL). PCR products were heated for 10 min at 957C
in an equal volume of denaturing loading buffer and
loaded onto an mutation-detection enhancement gel
(FMC). The gels were then dried and autoradiographed
for 2472 h. PCR products that showed an abnormal
SSCP pattern were directly sequenced on both strands
by the fluorometric method (DyeDeoxy or BigDye Ter-
minator Cycle sequencing kit; Applied Biosystems).
When possible, mutations were confirmed by restriction-
enzyme digestion on PCR amplification products.
In Situ Hybridization
Human embryos were collected from legally termi-
nated pregnancies, in agreement with French law and
the recommendations of the French Ethics Committee.
Embryos were staged under a microscope according to
the Carnegie classification criteria. Brain and organs of
a 20-wk fetus were also studied. Tissues were fixed with
4% paraformaldehyde, embedded in paraffin blocks,
and sectioned at 5 mm. PCR amplification of human
genomic DNA with SOX10 exon 5 primers 5/2F and
5/2R (GGTAATGTCCAACATGGAGACC and GTA-
GGCGATCTGTGAGGTGG, respectively) were cloned
into pCR-Script Amp SK(1) (Stratagene). Sense and
RNA polymerase in the presence of a35S-uridine
50(a-thio)triphosphate (1,200 Ci/mmol; NEN). Labeled
probes were purified on Sephadex G50 columns. Hy-
bridization and posthybridization washes were carried
out according to standard protocols (Wilkinson 1992).
Slides were dehydrated, exposed to Biomax MR X-ray
films (Amersham) for 3 d, dipped in Kodak NTB2 emul-
sion for 3 wk at 147C, and then developed and coun-
terstained with toluidine blue, coverslipped with Eukitt,
and analyzed with dark- and bright-field illumination.
No hybridization signal was detected with the a35S-la-
beled sense probes, which confirmed that the in situ hy-
bridization patterns of the a35S-labeled antisense probes
were specific. No cross-hybridization with SOX9 m-
RNAs was observed (data not shown).
Results
DNA Analysis
Abnormal SSCP patterns were detected in the SOX10
gene in all three WS4 patients studied. In each patient,
a heterozygous mutation was located in SOX10 exon
5, either at nucleotide 939 (TACrTAA) or at nucleotide
752 (TCGrTAG), which predicted nonsense mutations
at codon 313 (Y313X) and 251 (S251X), respectively
(fig. 1A, 1B). The Y313X mutation abolished a DdeI
restriction site and was found in families 1 (fig. 1C) and
2. In family 1, the mutation was inherited from the un-
affected mother (fig. 1C), whereas, in family 2, the mu-
tation was not inherited from the mother and the father
Touraine et al.: SOX10 Mutation in a Neurological Variant of WS4 1499

was not available for study. The S251X SOX10 trun-
cating mutation arose de novo in a boy (family 3) with
hypotonia and arthrogryposis who died soon after birth.
No other variation in the SOX10 gene was found, and
no abnormal SSCP pattern could be detected in the RET
gene-coding sequence in the patients studied. The pe-
culiar neurological phenotype, together with the similar
localization of SOX10 mutations in these WS4 patients,
prompted us to look for SOX10 expression during early
human brain development.
SOX10 Expression in Developing Humans
Table 2 summarizes the expression pattern of SOX10
in five human embryos in Carnegie stages 1317 (days
2843). At Carnegie 13, SOX10 was strongly expressed
in dorsal-root ganglia, otic vesicle (fig. 2A, 2B), and cells
migrating around the aorta, in branchial arches, and
toward the gut and tracheal bud (data not shown). At
Carnegie 16, signal intensity increased in each of these
areas (fig. 2C, 2D). In addition, nerve fibers originating
from the anterior horn of the spinal cord expressed
SOX10 over their whole length up to the target tissues,
whereas the anterior horn was negative (fig. 2D). At
Carnegie 17, SOX10 expression increased in ganglia and
extended to cranial ectomesenchyme, inner ear (semi-
circular canals, utricle, and saccula), and some regions
of the maxilla, mandible, and tongue, consistent with
the labeling of innervating fibers (fig. 2F). In the gut,
expression was restricted to the outer muscular layer,
where the enteric plexuses are located. Testis and de-
veloping pancreas were also SOX10 positive (data not
shown). No SOX10 transcripts could be detected in the
central nervous system at any of the embryonic stages
studied.
No SOX10 expression in the cerebral cortex and brain
nuclei was observed consistently in a 20-wk fetal brain.
However, a diffuse but variable expression was seen in
the other areas of the brain. High-power magnification
showed that the signals were located in some strongly
positive nonneuronal cells scattered in several glial-rich
areas, namely pons (fig. 3A, 3B), cerebellum (fig. 3H),
tectum of peduncle (fig. 3E), internal capsule (fig. 3D),
and hippocampal formations (fig. 3F, 3G), consistent
with SOX10 expression in glial cells. Fetal expression
of SOX10 was also observed in the enteric and tracheal
nervous systems, in pancreas acini, and in adrenal me-
dulla (data not shown). Finally, no expression was de-
tected in heart, kidney, thymus, spleen, or muscle. These
results were in agreement with previous data (Bondur-
and et al. 1998).
Discussion
WS4 is a rare neurocristopathy that results in the absence
of melanocytes and inner ear cells (WS) and parasym-
pathetic enteric neurons of the terminal hindgut (HSCR;
McKusick 1973; Branski et al. 1979; Omenn and
McKusick 1979; Shah et al. 1981). WS4 is transmitted
either as a recessive or a dominant autosomal trait, de-
pending on whether the endothelin-signaling pathway
(Puffenberger et al. 1994; Attie et al. 1995; Edery et al.
1996; Bidaud et al. 1997) or SOX10 mutations (Pingault
et al. 1998) are involved, respectively. Yet several WS4
cases are not related to EDNRB, EDN3, or SOX10

Table 2
SOX10 Expression in Early Human Embryogenesis
EXPRESSION AT CARNEGIE STAGEa
VARIABLE 13 14 16 17
Central nervous system 2 2 2 2
Peripheral nervous system:
Enteric nervous system Migrating cells 1 11 111
Tracheal nervous system Migrating cells 1 11 111
Ganglia:
Cranial 1 1 11 111
Dorsal root 1 1 11 111
Sympathetic 1 1 11 111
Peripheric nerves 2 2 11 111
Ectomesenchyme 2 1 11 Meckel 11; nasal
cartilage 111
Otic vesicle 1 1 11 Internal ear 111
Testis 2 2 2 111
Pancreas 2 2 2 111
Lingual glands 2 2 2 111
NOTE.A plus sign (1) indicates presence, and a minus sign (2) indicates absence
of the indicated expression. 1, 11, and 111 = relative levels of expression.
a
Carnegie stage corresponds to the following ages: Carnegie 13, 2831 d; Carnegie
14, 32 d; Carnegie 16, 3740 d; and Carnegie 17, 4143 d.
1500 Am. J. Hum. Genet. 66:14961503, 2000

Figure 2 SOX10 expression in transverse sections of a Carnegie 13 human embryo, under dark-field illumination. A, Cranial nerve ganglia
(V, VII/VIII, and IX/X) and otic vesicle (OV). Also shown, neural tube (NT). B, Dorsal root ganglia (DRG). SOX10 expression in transverse
sections through a Carnegie 16 human embryo under dark-field illumination. Also shown, heart (H) C, Cranial nerve ganglia V and ectome-
senchyme (EM). Also shown, nasal pits (NP), optic vesicle (Opv), rhombencephalon (Rh), and Rathke pouch (RP). D, Dorsal root (DR), dorsal
root ganglia (DRG), spinal nerve (SN), and ventral root (VR) merging from the anterior horn (AH). Serial parasagittal sections through a
Carnegie 17 human embryo, either hematoxylin and eosinstained (E) or hybridized with SOX10 riboprobe and examined under dark-field
illumination (F), which shows SOX10 expression in the cranial ganglia (V, VII/VIII, and IX/X), olfactory bulb (OlfB), cranial ectomesenchyme
(i.e., Meckel cartilage [MC] and nasal cartilage [NC]), nerves in intercostal spaces, nerve fibers (N) in maxilla (Max), mandible (Man), and
tongue (To), internal ear (semicircular canals [SCC], utricle, and saccula [U/S]), and the ultimobranchial body (UBB).

mutations. Among a total of 12 probands who presented
with WS4 features (deafness, pigment anomalies, and
HSCR), we have studied members of three families with
a progressive neurological phenotype. There have been
very few reports of WS patients with neurological in-
volvement, including mental retardation with seizures or
dystonia, muscular stiffness, and peripheral neuropathy,
but none had HSCR (Kawabata et al. 1987; Cantani et
al. 1989). Conversely, only two siblings have been re-
ported with extensive digestive dysganglionosis, ataxia,
peripheral neuropathy, and dysautonomic features but
no feature of WS (Schuffler et al. 1978). Therefore, when
we started this study, the patients presented an as-yet-
unreported association. Although several hypotheses re-
garding the additional neurological symptoms could
have been raised (such as a contiguous gene syndrome),
we ascribed the disorder to truncating SOX10 mutations
in individuals in each of the three families studied. This
hypothesis was further supported by two articles that
describe WS4 patients with SOX10 mutation and neu-
rological features, published while this article was in
preparation. Two siblings who harbored phenotypes
very similar to those of the patients we studied, including
nystagmus and ataxic cerebral palsy (although only one
of them had HSCR), were found to be heterozygous for
the Q377X SOX10 mutation (Southard-Smith et al.
1999). Furthermore, Inoue et al. (1999) described a pa-
tient with WS4, demyelinating neuropathy, and severe
leukodystrophy. This patient harbored a heterozygous
1400del12 SOX10 mutation that modified the 467 stop
codon and extended the SOX10 protein with 82 amino
acids.
SOX10 is a member of the SOX family of transcrip-
tion factors and local organizers of chromatin structure
(Pevny and Lovell-Badge 1997). Binding of SOX-pro-
tein high-mobility group (HMG) domain to the specific
DNA sequence motif A/T A/T CAA A/T G results in
DNA bending. There is no evidence for target specificity,
but rather there is evidence for tissue- and stage-specific
expression, which delimits a specific role for each of the
SOX family members. We speculate that the SOX10
Y313X and S251X nonsense mutations are responsible
for both WS4 and the additional neurological pheno-
type. Indeed, our data about SOX10 expression support
Touraine et al.: SOX10 Mutation in a Neurological Variant of WS4 1501

Figure 3 SOX10 expression in fetal brain. Transverse sections of the brain of a 20-wk fetus. A, F, H, hematoxylin and eosin staining.
B, C, D, E, G, I, SOX10 in situ hybridization under dark-field illumination. C is a higher-magnification view of B (boxed area). A diffuse
SOX10 expression is seen in many areas of the brain, located in some strongly positive cells scattered in the white matter, respecting the
pyramidal tracts (A, B). Signal is more intense in pons (AC); internal capsule (D); peduncle, especially in tectum (E); hippocampal formation
(F, G); and cerebellum (H, I). J, Adjacent slide to I, hybridized with the sense (control) SOX10 probe. Cranial nerves are strongly labeled as
shown for the nerve of cranial nerve ganglia VIII N (B, C). Py = pyramidal tracts.

its important role in neural-crest and brain develop- 4; Kuhlbrodt et al. 1998b; Pingault et al. 1998). Since
ment, especially since autonomous nervous system ab- the patients with WS4 reported here also carry hetero-
normalities are involved in the patientsnamely, re- zygous mutations that truncate the last 216 and 154
duction in sweat, tear, and saliva production. Both the amino acids residues of SOX10 for the S251X and
distribution and the morphology of the strongly positive Y313X mutations, respectively, there is no clear expla-
cells scattered in fetal-brain white matter support the nation for the phenotypic differences. However, it is
expression of SOX10 in glial cellsnamely, oligoden- worth noting that, so far, only the S251X and Y313X
drocytes, astrocytes, or immature glial cellsrather mutations are located between the HMG box and the
than neurons (fig. 3). Interestingly, in rodents, Sox10 is
expressed in glial precursors and then in oligodendro-
cytes (Kuhlbrodt et al. 1998a). Although it is premature
to establish the involvement of a specific subtype of glial
cells in the central nervous system symptoms observed
in the patients we studied, we can speculate on astrocyte
impairment, according to recent data, which shows that
cerebellar ataxia developed in mice after postnatal de-
struction of astrocytes (Delaney et al. 1996), whereas
oligodendrocyte dysfunction would instead result in
leukodystrophy.
The spectrum of SOX10 mutations reported to date
supports the hypothesis that haploinsufficiency is the Figure 4 Representation of SOX10 protein with the transacti-
mechanism for dominance in WS4. Indeed, when we vation domain and the HMG box, which shows the position of the
seven previously reported mutations, the mutation responsible for the
used a functional assay for SOX10 transcriptional ac- murine Dom mutant, and the two novel mutations. The mutations
tivity, two nonsense and one frameshift SOX10 muta- associated with an additional neurological phenotype are provided
tions resulted in loss of function, as shown in vitro (fig. (upper part).
1502
transactivation domain, in a SOX10 region of unknown
function (fig. 4). In addition, patients with SOX10 mu-
tations that result in either more proximal or more distal
truncations, such as E189X, Y207X, or 1076delGA (in
order, from more proximal to more distal), did not show
any neurological symptoms (Pingault et al. 1998; Sou-
thard-Smith et al. 1999). The patients with nystagmus
and ataxic cerebral palsy, briefly described by Southard-
Smith et al. (1999), had a truncating mutation, Q377X,
that removed only the transactivation domain. Con-
versely, the patient described by Inoue et al. (1999) had
a mutation at the SOX10 30 end that resulted in the
addition of 82 amino acids to the protein. Nevertheless,
a possible hypothesis for the extended clinical spectrum
in these patients would be that the S251X, Y313X,
Q377X, and 1400del12 SOX10 mutations resulted in
a dominant negative effect. Indeed, such proteins with
either an absent or a drastically modified transactivation
domain, but that retained the HMG box, might have
an enhanced DNA binding ability, blocking the prom-
otor of SOX10-responding genes, without transacti-
vation. Similar mechanisms have been demonstrated for
other mutant transcription factors such as PAX6 (Singh
et al. 1998) and MGF-Stat5 (Moriggl et al. 1996). Un-
fortunately, we failed to amplify by reverse transcrip-
tasePCR any SOX10 mRNA, either normal or mutant,
from lymphoblasts or leukocytes from patient 2 (data
not shown). Therefore, mutant mRNA instability can-
not be excluded: such a phenomenon, which has been
reported several times in various genes, would preclude
any dominant negative effect (Maquat 1995). Addi-
tional patients with a SOX10 mutation are necessary
to confirm the preliminary correlation between the phe-
notype and the location of truncating mutation. Alter-
natively, the wide phenotypic spectrum of SOX10 mu-
tations may be explained by the effect of modifier genes,
such as those identified in oligogenic neurocristopathies
such as nonsyndromic HSCR (Puffenberger et al. 1994;
Salomon et al. 1996; Doray et al. 1998; Bolk et al.
2000), especially since the reported cases have been as-
certained through the intestinal phenotype. The absence
of HSCR in one of the two sibs with nystagmus and
ataxic cerebral palsy, ascribed to the Q377X SOX10
mutation (Southard-Smith et al. 1999), supports this
hypothesis, but this is not sufficient to exclude the other
mechanisms as far as the neurological phenotype is
concerned.
Our study also emphasizes the apparently high rate
of new mutations at the SOX10 locus. Indeed, in a total
of 11 informative WS4 families studied, 7 of 11 SOX10
mutations arose de novo in the probands (Pingault et
al. 1998; Inoue et al. 1999; Southard-Smith et al. 1999;
this report; unpublished data). This does not seem to
be related to a mutational hot spot, since most of the
SOX10 mutations are different. Along this line, there
Am. J. Hum. Genet. 66:14961503, 2000
is no clear explanation for the recurrence of the SOX10
Y313X mutation, since it does not involve a CpG di-
nucleotide. It is thus possible that the high proportion
of new mutations is due to an observation bias related
to the severity of the WS4 phenotype, since most af-
fected individuals are unlikely to reproduce. Of interest,
in family 1, the mother of the proband did not show
any symptoms, although she carried this Y313X mu-
tation. It can be attributable either to a lack of pene-
trancewhich would be surprising as far as a dominant
negative effect of the mutation is postulatedor to so-
matic mosaicism. Indeed, the mutated Y313X allele in
the mother had a significantly weaker intensity than the
normal allele, compared with the pattern of her affected
son (fig. 1). Furthermore, the DNA sequence of SOX10
exon 5 in the mother was almost normal, with the mu-
tant peak barely visible on the electropherogram (data
not shown). Altogether, this favored the possibility of
mosaicism. Unfortunately, the mother refused to allow
us to further investigate her tissue samples. Mosaicism
seemed to be present also in the father of the family
described by Southard-Smith et al. (1999), as associated
with the Q377X SOX10 mutation.
This study highlights the role of SOX10 as an im-
portant factor in development of both neural-crest
derived cells and glial cells of the central nervous system,
which supports its involvement not only in WS4 but
also in an extended progressive neurological phenotype,
including growth retardation, impairment of the auton-
omous system (asialia, alacrima, and hypohidrosis), and
central nervous system anomalies, namely neonatal hy-
potonia with arthrogryposis, ataxia, pyramidal signs,
nystagmus, seizure, and mental retardation. Thus, these
data suggest that SOX10 might be regarded as an in-
teresting candidate gene in other neurodegenerative dis-
orders, especially since the penetrance for HSCR is in-
complete in WS4.
Acknowledgments
We thank Dr. Peter H. Byers for interesting comments and
kind attention. This study was supported by grants from the
Association Francaise contre les Myopathies, the Ligue contre
le Cancer (Comite de Paris), and the Association pour la Re-
cherche contre le Cancer. R.L.T. is a fellow from the Societe
dEtudes et de Soins pour les Enfants Paralyses et Polymal-
formes and the Fondation des Treilles.
Electronic-Database Information
Accession numbers and URL for data in this article are as
follows:
Online Mendelian Inheritance in Man (OMIM), http://www
.ncbi.nlm.nih.gov/Omim (for WS [MIM 193500, MIM
Touraine et al.: SOX10 Mutation in a Neurological Variant of WS4
193510, MIM 600193, and MIM 148820], HSCR [MIM
142623], and WS4 [MIM 277580])
References
Attie T, Till M, Pelet A, Amiel J, Edery P, Boutrand L, Munnich
M, et al (1995) Mutation of the endothelin-receptor B gene
in the Waardenburg-Hirschsprung disease. Hum Mol Genet
4:24072409
Bidaud C, Salomon R, Van Camp G, Pelet A, Attie T, Eng C,
Bonduelle M, et al (1997) Endothelin-3 gene mutations in
isolated and syndromic Hirschsprung disease. Eur J Hum
Genet 5:247251
Bolk S, Pelet A, Hofstra RMW, Angrist M, Salomon R,
Croaker D, Buys CHCM, et al (2000) A human model for
multigenic inheritance: phenotypic expression in Hirschs-
prung disease requires both the RET gene and a new 9q31
locus. Proc Natl Acad Sci USA 97:268273
Bondurand N, Kobetz A, Pingault V, Lemort N, Encha-Razavi
F, Couly G, Goerich DE, et al (1998) Expression of the
SOX10 gene during human development. FEBS Lett 432:
168172
Branski D, Dennis NR, Neale JM, Brooks LJ (1979) Hirschs-
prungs disease and Waardenburgs syndrome. Pediatrics 63:
803805
Cantani A, Bamonte G, Tacconi ML (1989) Mental retardation
and EEG abnormalities in Waardenburgs syndrome: two
case reports (EEG anomalies in Waardenburgs syndrome).
Padiatr Padol 24:137140
Delaney CL, Brenner M, Messing A (1996) Conditional ab-
lation of cerebellar astrocytes in postnatal transgenic mice.
J Neurosci 16:69086918
Doray B, Salomon R, Amiel J, Pelet A, Touraine R, Billaud
M, Attie T, et al (1998) Mutation of the RET ligand, Neur-
turin, supports multigenic inheritance in Hirschsprung dis-
ease. Hum Mol Genet 7:14491452
Edery P, Attie T, Amiel J, Pelet A, Eng C, Hofstra RM, Martelli
H, et al (1996) Mutation of the endothelin-3 gene in the
Waardenburg-Hirschsprung disease (Shah-Waardenburg
syndrome). Nat Genet 12:442444
Inoue K, Tanabe Y, Lupski JR (1999) Myelin deficiencies in
both the central and the peripheral nervous systems asso-
ciated with a SOX10 mutation. Ann Neurol 46:313318
Kawabata E, Ohba N, Nakamura A, Izumo S, Osame M
(1987) Waardenburg syndrome: a variant with neurological
involvement. Ophthalmic Paediatr Genet 8:165170
Kuhlbrodt K, Herbarth B, Sock E, Hermans-Borgmeyer I, Weg-
ner M (1998a) Sox10, a novel transcriptional modulator in
glial cells. J Neurosci 18:237250
Kuhlbrodt K, Schmidt C, Sock E, Pingault V, Bondurand N,
1503
Goossens M, Wegner M (1998b) Functional analysis of
sox10 mutations found in human Waardenburg-Hirschs-
prung patients. J Biol Chem 273:2303323038
Maquat LE (1995) When cells stop making sense: effects of
nonsense codons on RNA metabolism in vertebrate cells.
RNA 1:453465
McKusick VA (1973) Congenital deafness and Hirschsprungs
disease. N Engl J Med 288:691
Moriggl R, Gouilleux-Gruart V, Jahne R, Berchtold S, Gart-
mann C, Liu X, Hennighausen L, et al (1996) Deletion of
the carboxy-terminal transactivation domain of MGF-Stat5
results in sustained DNA binding and a dominant negative
phenotype. Mol Cell Biol 16:56915700
Omenn GS, McKusick VA (1979) The association of Waar-
denburg syndrome and Hirschsprung megacolon. Am J Med
Genet 3:217223
Pevny LH, Lovell-Badge R (1997) Sox genes find their feet.
Curr Opin Genet Dev 7:338344
Pingault V, Bondurand N, Kuhlbrodt K, Goerich DE, Prehu
MO, Puliti A, Herbarth B, et al (1998) SOX10 mutations
in patients with Waardenburg-Hirschsprung disease. Nat
Genet 18:171173
Puffenberger EG, Hosoda K, Washington SS, Nakao K, deWit
D, Yanagisawa M, Chakravarti A (1994) A missense mu-
tation of the endothelin-B receptor gene in multigenic
Hirschsprungs disease. Cell 79:12571266
Read AP, Newton VE (1997) Waardenburg syndrome. J Med
Genet 34:656665
Salomon R, Attie T, Pelet A, Bidaud C, Eng C, Amiel J, Sar-
nacki S, et al (1996) Germline mutations of the RET ligand,
GDNF, are not sufficient to cause Hirschsprung disease. Nat
Genet 14:345347
Schuffler MD, Bird TD, Sumi SM, Cook A (1978) A familial
neuronal disease presenting as intestinal pseudoobstruction.
Gastroenterology 75:889898
Shah KN, Dalal SJ, Desai MP, Sheth PN, Joshi NC, Ambani
LM (1981) White forelock, pigmentary disorder of irides,
and long segment Hirschsprung disease: possible variant of
Waardenburg syndrome. J Pediatr 99:432435
Singh S, Tang HK, Lee J-Y, Saunders GF (1998) Truncation
mutations in the transactivation region of PAX6 result in
dominant-negative mutants. J Biol Chem 273:2153121541
Southard-Smith EM, Angrist M, Ellison, JS, Agarwala RR,
Baxevanis AD, Chakravarti A, Pavan WJ (1999) The
Sox10Dom mouse: modeling the genetic variation of Waar-
denburg-Shah (WS4) syndrome. Genome Res 9:215225
Wilkinson DG (1992) In situ hybridization. In: Rickwood D,
Hames BD (eds) The practical approach series. Vol 109. IRL
Press, Oxford